Leading Teams: A Data Scientist’s Perspective

Can you remember some of the leaders that inspired you? What did they have in common? Think about this for a minute before reading on.

The previous post of this series focused on the technical details of managing people. This article will cover what it means to lead a team, and some tips I would give to an aspiring leader.

Just as I did in the previous article, leadership is a skill I am continuously improving. I have definitely not figured it out. Having been promoted to a Team Lead position three years ago, I still remember what it feels like to start.

Throughout my learning journey, I found the books: High Output Management by Andy Grove, Turn the Ship Around! by David Marquet and Extreme Ownership by Jocko Willink and Leif Babin particularly helpful.

This article summarises some of the things I would have liked to be told when I started as a Team Lead. This is a personal take, I hope you find some of them useful.

Take responsibility for everything

This was the real turning point of my leadership experience. The moment I realised that the performance of the team was the responsibility of the team lead. Managers cannot blame individual employees for poor team performance.

The team will perform poorly if you did not do the following properly:

• Set the right quality standards
• Communicated goals and priorities early enough
• Defined tasks clearly enough
• Put the right guardrails in place
• Planned for delays or emergencies
• Trained your team to do the job
• Selected the right team members for the job

These are just a few examples to realise that managers and leaders must be fully responsible for the performance of their team. Poor performance can always be avoided, or at least largely mitigated by excellent leadership.

With this mindset, I started seeing every failure or setback as a way to improve my team management skills and the way in which I do my job.

Goal alignment

We all work to achieve something. We all have goals. These could include:

• Becoming a Data Science expert
• Building highly accurate models
• Making the world a better place
• Making as much money as possible
• Getting a fancy job title at a good company as a status symbol
• etc…

The key to a team’s success is to align these goals. To do so, I started making the goals of our team clear:

• Flawlessly maintain a critical production Machine Learning service
• Deliver business impact through Machine Learning model improvements
• Communicate all of this with business stakeholders

This way, everyone in the team knows what the objectives are. They also know that good performance in the above three will be rewarded by what they are looking for. As a manager and leader, the key is to link team members’ personal goals to success in the team’s main objectives.

Foster ownership

Give people the ownership of their work. As a beginning manager, I had a tendency to be very involved in both the delivery and communication of our team’s work. My work life changed when I started giving ownership to my team members.

A new open-ended problem with no clear solution? A presentation to a senior stakeholder? The exploration of a new modelling approach? All of these are opportunities to give ownership to the team member who will do the work.

As a side note, this also makes life as a manager much easier.

Prioritisation is key

In knowledge work, time and effort are usually the main bottlenecks. There are so many things to work on, what should your team focus on? As a team lead, you are responsible for this effort allocation. It is up to you to determine the tasks that will bring you closest to your team’s goals.

Even if the team lead is eventually responsible, I like involving team members in prioritisation discussions.

Prioritisation is critical for both your team’s time and your own time. The key is to identify the tasks that have the highest impact. Try to eliminate or automate the rest.

As a manager, I find that some of the most impactful uses of my time are:

• Spending an hour of free thought and research on a tricky topic
• Feedback and project talks with my team members
• Preparation for meetings with senior management

This last part is inspired by Andy Grove’s majestic High Output Management

Taking the team’s perspective, prioritisation generally comes down to the balance of short-term ad hoc tasks and long-term projects. Any advice I could give now would be too general. My overall recommendation would be to consider effort and expected business impact as key criteria - maybe including risk too. See the RICE Model as way to deal with these trade-offs.

Ruthless organisation

The team lead is responsible for the consistent delivery of work. I take it upon myself to take note of everything, set up reminders, or keep track of work in progress. The team lead’s role is to manage a constant flow of information. Information is coming both from above, stakeholders and management; and below, with updates on the tasks in progress.

I would describe this aspect of my role as: information sponge, connector of dots, note taker in chief and reminder meister. These titles are not as sexy as “team lead” or “manager”, but a good description of some of the day-to-day activities of leading teams.

Final Thoughts

Summarising this blog post in one sentence: keep the goal in mind, make sure this goal is clear and shared, sort out all of the details to get there - it’s all up to you.

Leadership is a skill I am constantly trying to improve. I look forward to revisiting this article as I learn more along the way. If you have found some of it useful, subscribe to my newsletter for more articles like this.